Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261238AbUKWNv4 (ORCPT ); Tue, 23 Nov 2004 08:51:56 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261246AbUKWNv4 (ORCPT ); Tue, 23 Nov 2004 08:51:56 -0500 Received: from mx2.elte.hu ([157.181.151.9]:28605 "EHLO mx2.elte.hu") by vger.kernel.org with ESMTP id S261243AbUKWNvu (ORCPT ); Tue, 23 Nov 2004 08:51:50 -0500 Date: Tue, 23 Nov 2004 15:53:45 +0100 From: Ingo Molnar To: Rui Nuno Capela Cc: Florian Schmidt , linux-kernel@vger.kernel.org, Lee Revell , mark_h_johnson@raytheon.com, "K.R. Foley" , Bill Huey , Adam Heath , Thomas Gleixner , Michal Schmidt , Fernando Pablo Lopez-Lezcano , Karsten Wiese , Gunther Persoons , emann@mrv.com, Shane Shrybman , Amit Shah , Esben Nielsen Subject: Re: [patch] Real-Time Preemption, -RT-2.6.10-rc2-mm2-V0.7.30-2 Message-ID: <20041123145345.GA21429@elte.hu> References: <20041122005411.GA19363@elte.hu> <20041122020741.5d69f8bf@mango.fruits.de> <20041122094602.GA6817@elte.hu> <56781.195.245.190.93.1101119801.squirrel@195.245.190.93> <20041122132459.GB19577@elte.hu> <20041122142744.0a29aceb@mango.fruits.de> <65529.195.245.190.94.1101133129.squirrel@195.245.190.94> <20041122154516.GC2036@elte.hu> <9182.195.245.190.93.1101142412.squirrel@195.245.190.93> <20041123144622.GA20085@elte.hu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20041123144622.GA20085@elte.hu> User-Agent: Mutt/1.4.1i X-ELTE-SpamVersion: MailScanner 4.31.6-itk1 (ELTE 1.2) SpamAssassin 2.63 ClamAV 0.73 X-ELTE-VirusStatus: clean X-ELTE-SpamCheck: no X-ELTE-SpamCheck-Details: score=-4.9, required 5.9, autolearn=not spam, BAYES_00 -4.90 X-ELTE-SpamLevel: X-ELTE-SpamScore: -4 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1095 Lines: 24 > since the client runs as SCHED_OTHER, doesnt this mean that simple > delays between SCHED_OTHER tasks could cause xruns in jackd too? A > SCHED_OTHER task can be delayed indefinitely at any stage. So shouldnt > the test-clients have RT priority as well, to guarantee xrun-less > jackd? if SCHED_OTHER is the goal, then the xruns you are seeing in the fluidsynth test are likely the result of random fluctuations in scheduling of SCHED_OTHER tasks. The way to find out whether that's the main source of xruns would be the following test: start each fluidsynth instance as "nice -19 fluidsynth ..." to renice it to +19. Nice +19 gives the smoothest timeslices to CPU-bound tasks (such as fluidsynth), and should give the smallest global fluctuations. (The system must be idle during the test of course, a nice +19 task is easily preempted.) Ingo - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/